Supramolecular organization and magnetic properties of mesogen-hybridized mixed-valent manganese single molecule magnets [MnIII 8MnIV4O12(L x, y, z -CB)16(H2O)4]

Article

pubs.acs.org/JACS

Supramolecular Organization and Magnetic Properties of Mesogen-

Hybridized Mixed-Valent Manganese Single Molecule Magnets

[Mn^IIIMn^IVO₁₂(L_x,y,z‑CB)₁₆(H₂O)₄]

8

4

Emmanuel Terazzi,^†Guillaume Rogez, Jean-Louis Gallani, and Bertrand Donnio*

́ ́

Institut de Physique et Chimie des Materiaux de Strasbourg (IPCMS), UMR 7504, CNRS-Universite de Strasbourg, 23 rue du Loess,

BP43, 67034, Strasbourg cedex 2, France

S

* Supporting Information

ABSTRACT: Single molecule magnets (SMM) may be

considered for the construction of future integrated nano-

devices, provided however that some degree of ordering is

imparted to these molecules (surfaces nanostructuration).

Combining such nanoobjects with liquid-crystalline orderings

to control their assembly and to potentially address them

individually therefore appears as one promising strategy. Four

mesomorphic, mixed-valent [Mn^III₈Mn^IV₄O₁₂(L_x,y,z‑CB

)

16

(H₂O)₄] SMM, diﬀering in the number of liquid-crystalline

promoters, (L_x,y,z‑CB), were synthesized, and their self-

organizing and magnetic properties were investigated. The inﬂuence of the peripheral modiﬁcations, and precisely how

supramolecular ordering and magnetic properties may be aﬀected by the evolution of the proto-mesogenic cyanobiphenyl-based

ligands substitution pattern, was explored. Small-angle X-ray scattering studies revealed that all of the hybridized clusters self-

organize into room-temperature bilayer smectic phases, mandated by the speciﬁc mesogenic functionalization and that the

polymetallic cores are further organized according to a short-range pseudo-2D lattice with hexagonal and/or square symmetry.

All mesomorphous hybridized dodecamanganese complexes still behave as SMM: they exhibit blocking of the magnetization at

about 2.6 K as evidenced by the occurrence of frequency-dependent out-of-phase ac susceptibility signals as well as an opening of

the hysteresis cycle with coercive ﬁelds varying between 0.13 and 0.6 T, depending on the surface ligands topology. Comparison

of the magnetic properties within this series reveals intricate correlations between the structural features of the mesomorphous

molecule magnet (i.e., symmetry of the ligands substitution patterns, molecular conformation, average intercluster distances, and

respective inclination) with respect to the relative proportion of slow- and fast-relaxing species and the absolute values of the

coercive ﬁelds.

quantum algorithms in computers because of their intrinsic

quantum nature.^3,4High-density information storage, molecular

spintronics, and processing at the molecular level could indeed

be the technological jump leading to the next generation of

computers. Still, a good number of progresses and break-

throughs remain to be achieved before any of these dreams

come true: rational design and organization of SMM on a

surface remain elusive,⁵addressing a single molecule is a

technological challenge,⁶and all of the SMM discovered until

now lose their properties above 6 K at best.⁷

SMM usually possess a high-spin ground state brought by 3d

(and also 4d and/or 4f)^1,7metal ions: the highest spin value

observed so far is S = 51/2 for a SMM Mn25 complex⁸(the

highest spin reported for a non-SMM compound is S = 83/2

for a Mn19 cluster⁹). The ﬁrst and now very famous SMM

cluster is the so-called mixed-valent “manganese 12” (Mn12)

with the general formula [Mn^III₈Mn^IV₄O₁₂(CH₃CO₂)₁₆(H₂O)₄]

(Figure 1), [Mn₁₂(OAc)₁₆] for short, serendipitously synthe-

INTRODUCTION

■

Usually, for traditional materials, ferromagnetism is a collective

phenomenon and is the fact of an enormous number of atoms,

organized in such a way that their electrons’ magnetic moments

interact cooperatively to form a permanent magnet. The

discovery that some peculiar molecules, for example, the so-

called single molecule magnets (SMM), could also function

individually as single “nano”-magnets by retaining a magnet-

ization at the molecular level revived intense research activity in

the ﬁeld of molecular magnetism.¹Indeed, below a certain

temperature, T_B, known as the blocking temperature, and in the

absence of a magnetic ﬁeld, the magnetic moment of a SMM

stops ﬂuctuating and becomes stable, without the need for

domains, walls, and long-range ordering of the magnetic

moments (no exchange interaction). Independent of their

academic interest, SMM, among other nanomagnetic materi-

als,²may be considered as ideal candidates to transcend the

superparamagnetic frontier, which currently limits the technol-

ogy of magnetic information storage to densities of the order of

400 Gb/in.²(e.g., gigabit per square inch).³They moreover are

envisioned as building blocks (qubits) for implementing

Received: November 14, 2012

Published: January 22, 2013

2708

dx.doi.org/10.1021/ja311190a | J. Am. Chem. Soc. 2013, 135, 2708−2722

Journal of the American Chemical Society

Article

state stems from the magnetic anisotropy of the eight

outermost Mn^IIIions. The coordination around each Mn^III

ion is actually not a regular octahedral, but is Jahn−Teller (JT)

distorted, as two bonds are longer than the other four (JT

elongation).¹⁷⁻¹⁹These crystal-ﬁeld distortions together with a

spin−orbit interaction result in a zero-ﬁeld interaction at each

Mn^IIIion; hence, vectorial projection of single-ion anisotropies

on to the ground state gives rise to the easy axis anisotropy.

Numerous, and versatile, homo- and heterometallic molec-

ular structures (based on 3d−5d and/or 4f metal ions with all

sorts of peripheral ligands),^1,7,12from single metal-ion to high-

nuclearity complexes (e.g., Mn84²⁰), have been synthesized, as

well as various approaches (e.g., ligand surface-modiﬁcation,²¹

applying pressure^22,23) have been undertaken, which were

directly aiming at improving one or several of these parameters

(S, D, U_eff). To the chemist’s despair, however, the original

compound often remains the best compromise. Even though

sometimes successful, the random testing of new structures is

not the most eﬃcient way. The traditional approach therefore

usually relies on trying to modify one parameter at a time and

look for the subsequent evolution of the physical properties.

Unfortunately, the intramolecular magnetic couplings are

exquisitely sensitive to bonds’ lengths and angles, making the

accurate tweaking somehow hazardous: the nature of the

magnetic coupling between metal centers, and particularly

within Mn-based cores, highly depends on the mechanical

inﬂuence of peripheral ligands, because the magnetic coupling

are often mediated by O or N atoms and sensitive to bond

angles and interatomic distances (Mn12,^24,25Mn6^7,16,26). For

[Mn₁₂(OAc)₁₆], the correlation between magnetic properties

and molecular geometry was conﬁrmed because changes were

essentially ascribed to either variation of the angle between the

JT axis and the S₄molecular axis or enhancement of JT

isomerism. The inﬂuence of the torsion of the JT axis on the

magnetic anisotropy is also theoretically expected.^27,28All of

these results prove that the deliberate distortions and

deformations that can be inﬂicted to the molecular structure

without its destruction can indeed inﬂuence the ﬁnal magnetic

properties, but have a limited action range, and therefore this

strategy will never enable the rise of the energy barrier up to an

arbitrarily high value. Applying an anisotropic pressure by

chemical means, that is, through the use of appropriate

anisotropic ligands, is a priori more interesting, although such

a synthetic control also appears as a diﬃcult challenge.

Liquid crystals (LC) are outstanding examples of soft,

molecular self-assemblies that exquisitely combine order and

ﬂuidity, within components self-organizing into a wide diversity

of mobile and long-range ordered, periodic structures.^29,30

Furthermore, LC assemblies can easily form thin ﬁlms, which

are intrinsically defect-tolerant because positioning errors are

corrected automatically during the process of self-assembly.

Their dynamic nature, function-integration, and stimuli-

responsiveness abilities further render them more inevitable

for applications in modern technologies and particularly in

organic electronics.³¹The formation of ordered yet ﬂuid LC

mesophases may thus be sought as a ﬁrst step in the route to

controlling the assembly of single molecule magnets into low-

dimensional supramolecular orderings for their ultimate

incorporation within functional macroscopic devices (molec-

tronic-, spintronic-, magneto-optic-, magneto-electric-based

systems).^3,5,6We have already reported how Mn12 clusters

could be endowed with mesomorphic character, without losing

their original low-temperature magnetic properties.³²The

Figure 1. Crystallographic structure of [Mn₁₂(OAc)₁₆] complex along

the crystal c-axis (top) and b-axis (bottom). Color scheme: Mn^IV,

yellow; Mn^III, purple; O, red; C, black.

sized by Lis in 1980,¹⁰but whose magnetic bistability was

demonstrated only a decade later by Gatteschi et al.¹¹This

prototypical molecular nanomagnet¹²has one of the highest

blocking temperature (T_B≈ 3.7 K, for a characteristic relaxation

time τ = 100 s, see below), its spin ground state reaches a rather

high value (S = 10), and its coercive ﬁeld also is quite

impressive (μ₀H_c≈ 1 T at 2 K). This mixed-valence

polyoxometallate core contains 12 manganese ions, and its

structure consists of a roughly planar disk made up of a central

Mn^IV₄O₄cubane moiety (Mn^IV, S = 3/2) surrounded by a ring

of eight Mn^IIIions (S = 2), connected to the cube by eight μ₃-

O²⁻ions and four κ²-μ₂-acetate groups perpendicular to the

plane of the disk (two on each side, Figure 1). The peripheral

Mn^IIIcenters are further connected to each other by eight

equatorial and four axial κ²-μ₂-acetate groups; consequently, the

Mn^IIIions of the outer ring are alternatively doubly bridged to

one and singly bridged to two Mn^IVions (referred to as Mn^IIIof

type I or II, respectively). Generally, the magnetic couplings

between the core metal ions are almost never entirely

ferromagnetic: some ions are ferromagnetically coupled, some

antiferromagnetically, and the total spin of the cluster therefore

rarely is the bare sum of all of the individual spins of the ions.

In [Mn₁₂(OAc)₁₆], the outer Mn^IIIions are coupled

ferromagnetically, and so are the inner Mn^IVions, but the

two subsets are coupled antiferromagnetically.¹³

As just a high spin is not enough, there must also be some

Ising magnetic anisotropy (easy magnetic axis), i.e., a negative

zero-ﬁeld splitting parameter, D, which creates an energy

barrier, U_eff, to the thermal relaxation of the magnetization.¹⁴

This barrier is around 61 K for [Mn₁₂(OAc)₁₆],¹⁵while the

highest value reported so far for a SMM is 86.4 K.¹⁶In

[Mn₁₂(OAc)₁₆], most of the magnetic anisotropy of the ground

2709

dx.doi.org/10.1021/ja311190a | J. Am. Chem. Soc. 2013, 135, 2708−2722

Journal of the American Chemical Society

Article

Scheme 1. Schematic Representation of the Solvent-Mediated Ligand-Exchange Reaction and Chemical Structures of the

Cyanobiphenyl-Substituted Benzoic Acid Derivatives Used as Ligands (HL_x,y,z-CB)

induction of thermotropic smectic or cubic mesophases was

achieved through chemical exchange reaction of the acetates

with protomesogenic gallic acid-based derivatives. We therefore

come closer to the aforementioned goal of 3D-organizing

functional entities at the nanoscale. Addressing nano-objects in

3D is quite challenging, but ﬁrst XMCD results did cast doubts

about the stability of the magnetic behavior of SMM on

surfaces.³³Shortly after, nevertheless, it was shown that the

magnetism in itself was not altered, but that the observed

behavior most probably was the result of an unforeseen

photomagnetic eﬀect.^34,35Interestingly, this unexpected photo-

sensitivity of LC Mn12 derivatives could allow addressing the

magnetic state of the molecule using a combination of light and

magnetic ﬁeld and therefore facilitate the use of these SMM in a

molecular memory.

We shall describe our attempt for controlling the supra-

molecular mesomorphic organization and the magnetic

parameters of a complete set of liquid-crystalline Mn12 single

molecule magnets containing cyanobiphenyl peripheral groups.

The design, synthesis, and functionalization by ligand-exchange

method, and the characterization of their self-organization into

LC smectic phases, and their peculiar SMM properties will be

described. The present study aimed at understanding and

rationalizing the inﬂuences of speciﬁc structural variations

introduced in the molecular system, the evolution of the

substitution pattern of the ligands, their consequences on the

supramolecular organizations, and the impact of space-

structuring on the magnetic properties. In particular, we shall

show how the use of anisotropic pro-mesogenic ligands can

provide a means for ﬁnely tuning the molecular geometry and

the magnetic properties without fundamental alteration.

adaptable) interface all around the “anti” mesogenic poly-

metallic core: the organic sheath, made of polypedal

protomesogenic ligands, has thus been designed to compensate

the large available surface and to enhance microsegregation

auspicious to the formation of LC mesophases. The surface

alteration of the native cluster [Mn₁₂(OAc)₁₆] is achieved by

the complete exchange of the 16 acetates with suitable

benzoate-based ligands, a method extensively exploited to

modify at wish the outer shell of the cluster, preserving at the

same time the original single molecular magnetic behavior,^7,12

and recently shown eﬃcient in promoting mesomorphism.^32,38

Since these earlier reports, other related strategies based on

either covalent binding or ionic self-assembly (ISA) have

successfully been employed to induce LC properties in a priori

other nonmesogenic, three-dimensional hybrid molecular

structures such as decaadduct-fullerene³⁹derivatives, nano-

meter-sized organometallic⁴⁰and highly charged polyoxometal-

late⁴¹clusters, and, to some extent, nanoparticular systems,⁴²to

name a few.

Mesomorphic, dodecanuclear manganese oxocomplexes with

the general formula [Mn₁₂O₁₂(L_x,y,z‑CB)₁₆(H₂O)₄] were there-

fore directly obtained by the complete replacement of the 16

LC-inert acetate groups of the precursory [Mn₁₂(OAc)₁₆]

compound with several proto-mesogenic cyanobiphenyl-con-

taining benzoates, diﬀering in the substitution pattern at the

anchoring benzene ring, by solvent-mediated ligand-exchange

reaction (HL_x,y,z‑CB, Scheme 1). To ensure phase formation, the

anchoring part was further decoupled from the mesogenic end-

groups by undecylene spacers. [Mn₁₂(OAc)₁₆] was prepared in

situ from the easily accessible MnOAc₂and KMnO₄precursors,

according to the standard procedure,¹⁰and the HL_x,y,z‑CBacid

derivatives were prepared following literature protocols.^32,43

Synthesis and Characterization of Ligands HL_x,y,z‑CB

and Corresponding Dodecanuclear Clusters [Mn₁₂O₁₂-

(L_x,y,z‑CB)₁₆(H₂O)₄]. Acids HL_3,5‑CB, HL_3,4‑CB, and HL_4‑CBwere

prepared in two steps starting by the Williamson etheriﬁcation

of the precursory 3,5-dihydroxybenzoate, 3,4-dihydroxy-

benzoate, and 4-hydroxy-benzaldehyde, respectively, with 4′-

(11-bromo-undecyloxy)-biphenyl-4-carbonitrile in acetone fol-

lowing published procedures (HL_3,4,5‑CBwas reported pre-

RESULTS

■

The strategy developed to endow the archetypical Mn12 SMM

cluster with LC properties has been inspired and adapted from

a previously described methodology used to successfully force

mesomorphism in hexakis(methano) C₆₀-containing com-

pounds.^36,37The idea consists of limiting the unfavorable

eﬀects of the bulky Mn12-based metal-coordination polyhedron

spatial expansion by creating a ﬂuid and deformable (e.g.,

2710

dx.doi.org/10.1021/ja311190a | J. Am. Chem. Soc. 2013, 135, 2708−2722

Journal of the American Chemical Society

Article

a

Scheme 2

a

Synthetic conditions: (i) Br-(CH₂)₁₁-OH, PPh₃, DIAD, THF, 0 °C → room temperature; (ii) 3,4,5-trihydroxybenzoate/3,5-dihydroxybenzoate/

3,4-dihydroxybenzoate/4-hydroxy-benzaldehyde, K₂CO₃, KI_cat, acetone; (iii) KOH, EtOH/THF, H₂O, HCl 37%; (iv) H₂NSO₃H, NaClO₂, H₂O/

THF.

viously).^32,43The methyl esters MeL_3,5‑CBand MeL_3,4‑CBwere

hydrolyzed into their corresponding carboxylic acids, whereas

the intermediate aldehyde derivative (AldL_4‑CB) was directly

converted into its acidic counterpart by treatment with the

oxidizing H₂NSO₃H/NaClO₂mixture (Scheme 2).

likely preclude proper peaks attribution. Moreover, the

nonequivalency of the 16 binding sites (i.e., axial and equatorial

positions, vide supra) and the possible occurrence of several

positional (e.g., water molecules coordination mode) and

conformational isomers (e.g., as in the case of low-symmetrical

ligands such as HL_3,4‑CB) pose serious limitations. However,

because dipolar coupling is distance-dependent, the peripheral

protons of the cyanobiphenyl end-moieties will be less aﬀected

than those of the anchoring benzoate, and closer to the

paramagnetic cluster metallic core, this technique may be used

as a probe of their binding. Peripheral aromatic systems of the

cyanobiphenyl units are thus particularly well adapted to this

analysis: their signals are located in a chemical-shift range free

of other peaks, and moreover are weakly broadened and shifted

compared to those of the benzoate anchor. The NMR spectra

of HL_4‑CBand its corresponding [Mn₁₂O₁₂(L_4‑CB)₁₆(H₂O)₄]

complex, both selected as excellent representatives of this eﬀect,

were measured in CDCl₃and prove the binding of the

benzoates onto the cluster (Figure 2, spectra of the other

compounds are in the Supporting Information). The signals of

the two sets of protons H₁₋₂have totally disappeared, a

consequence of their grafting, and thus proximity to the

paramagnetic cluster. In contrast, the presence of sets of

protons H₃₋₆signals at same chemical shifts, although

broadened, conﬁrmed the grafting of the new ligands onto

the cluster and in addition the complete absence of free ligand.

Similar results were obtained for the other cyanobiphenyl-

containing systems (Supporting Information).

Mesomorphic Properties of the Dodecamanganese

Clusters. Liquid-crystalline mesophases were induced in the

four hybrids upon heating as revealed by POM, DSC, and

small-angle X-ray scattering (SAXS) techniques. The thermal

behavior, structural data, and mesophase characteristics are

collected in Table 1. As veriﬁed by thermogravimetry (TGA,

Supporting Information), all clusters possess an enhanced

thermal stability with respect to the precursory acetate parent

(which decomposes at ca. 35 °C, in air), but decomposition still

occurs at high temperatures in the mesophase before reaching

the isotropic liquid phase: the onset of a continuous weight-loss

is variable and depends on the sample structure, and occurs at

about 150 °C for [Mn₁₂O₁₂(L_4‑CB)₁₆(H₂O)₄] and at ca. 200 °C

for the three other cluster compounds; a weight loss greater

than 2% is nevertheless not observed below 250 °C. Thus,

providing that the compounds are not maintained for long

periods of time above 150 °C, temperature-dependent analyses

can be performed. POM observations carried out during the

ﬁrst heating are not very conclusive, and the presence of

MS and elemental analyses probed the purity of all of the

cyanobiphenyl-containing organic derivatives (Supporting

1

Information). H NMR conﬁrmed the expected dynamically

averaged C_2vmolecular symmetry of HL_3,5‑CBand HL_4‑CB, and

1

the C_Ssymmetry of HL_3,4‑CB, on the H NMR time scale. The

liquid-crystalline properties of cyanobiphenyl-containing li-

gands and related precursors were studied by polarized-light

optical microscopy (POM) and diﬀerential scanning calorim-

etry (DSC). This analysis revealed the formation of a

metastable nematic (monotropic N) mesophase for all of the

ligands, except for HL_4‑CBwhose mesophase is enantiotropic

(Supporting Information), in agreement with published results

on related compounds.⁴³The clearing temperatures of the

ligands decrease with increasing substitution, from 193 °C

(HL_4‑CB) to 145−150 °C (HL_3,5‑CBand HL_3,4‑CB) and 141 °C

for HL_3,4,5‑CB(Supporting Information). All of the acids are

stable above their clearing temperatures.

The covalent functionalization of the dodecanuclear clusters’

outer shell was achieved by ligand-exchange reaction between

the parent [Mn₁₂(OAc)₁₆] cluster with a large excess of the

pro-mesogenic acids HL_x,y,z‑CBmediated in a CH₂Cl₂/toluene

solvent mixture (Scheme 1). After several successive azeotropic

distillations of the produced CH₃CO₂H in toluene (to displace

favorably the equilibrium), the crude hybridized clusters

[Mn₁₂O₁₂(L_x,y,z‑CB)₁₆(H₂O)₄] were puriﬁed by gel exclusion

chromatography (GPC) and obtained as room-temperature

waxy solids after vacuum drying. The structural characterization

of the clusters was performed in both solid state and solution,

relying mostly on elemental analysis, found to be in good

agreement with the expected stoichiometries, and by IR

spectroscopy, whose full substitution of the 16 coordination

sites, as well as complete removal of ligands in excess, was

clearly established (Supporting Information). Once covalently

grafted onto the clusters, the characteristic stretching vibration

band of the carbonyl groups of the free benzoates disappeared;

this absence also indicates that no free ligand or acetate was

present.

Because the presence of the paramagnetic manganese ions,

NMR spectroscopy is a priori not well suited (i.e., non

quantitative). Indeed, the dipolar coupling between protons

and d-electrons induces fast relaxation times. Consequently,

serious broadening of the signals and possible unexpected shifts

2711

dx.doi.org/10.1021/ja311190a | J. Am. Chem. Soc. 2013, 135, 2708−2722

Journal of the American Chemical Society

Article

Table 1. Thermal Transitions (Onset Temperatures) and

Mesophase Characteristics for the Four LC SMM

Compounds

L in cluster

[Mn₁₂O₁₂(L)₁₆(H₂O)₄]

transition

temp (°C)

mesophase

b

a

parameters

L_3,4,5‑CB

G_Sm40.5 Sm 200 T = 80−160 °C

dec.

d = 41.9 0.1 Å

h_cluster= 23.6 1.0 Å

V

_mol= 34 000 500 Å³(δ ≈ 1 g

cm⁻³

A_mol= 810 A², a_mes= 33.8 Å²

G_Sm16.3 Sm 200 T = 40−140 °C

)

L_3,5‑CB

dec.

d = 43.5 0.1 Å

h_cluster= 20.2 2.0 Å

V_mol= 25 100 500 Å³(δ ≈

0.96 g cm⁻³

)

A_mol= 580 20 A²,

a_mes= 36.1 Å²

L_3,4‑CB

G_Sm17.0 Sm 200 T = 40−200 °C

dec.

d = 45.3 0.1 Å

h_cluster= 20.9 2.0 Å

V

_mol= 24 000 500 Å³(δ ≈ 1 g

cm⁻³

)

A_mol= 530 20 A²,

a_mes= 33.1 Å²

L_4‑CB

G_Sm17.2 Sm 150 T = 40−140 °C

dec.

d = 47.0 0.1 Å

h_cluster= 19.0 3.0 Å

1

Figure 2. H NMR spectra with numbering scheme for [Mn₁₂O₁₂-

V_mol= 14 000 300 Å³(δ ≈

(L_4‑CB)₁₆(H₂O)₄] (top) and HL_4‑CB(bottom) in CDCl₃at 298 K.

1.03 g cm⁻³

A_mol≈ 300 A², a_mes= 37.2 Å²

)

a

Transition temperatures given from the second DSC cycle.

Abbreviations: G_Sm, glassy smectic phase; Sm, smectic phase; dec,

decomposition. d, smectic periodicity; h_cluster, intercluster distance in

the smectic sublayer, measured by XRD; V_mol, molecular volume; δ,

density; A_mol, molecular cross-section area: A_mol= V_mol/d; a_mes

transverse cross-section area per mesogenic unit: a_mes= 2A_mol/n_mes

(n_mes: number of cyanobiphenyl subunits per cluster). All XRD were

performed during the second and third heating cycles.

nonhomogeneous areas complicated a proper analysis. As for

[Mn₁₂O₁₂(L_3,4,5‑CB)₁₆(H₂O)₄] previously described,³²virgin

samples needed to be heated ﬁrst, then annealed for 1/2 min

at high temperatures (e.g., 80−120 °C), and then cooled. After

this thermal treatment, POM observations of the three new

clusters revealed persistent, highly birefringent, and homoge-

neous optical textures (Supporting Information) in accordance

with liquid-crystalline mesophase formation at or near room

temperature, the ﬂuidity of which increases as the temperature

is raised. Phase assignment by this method appeared however

adventurous, as no speciﬁc natural texture could be obtained

because all of the clusters degrade before reaching the isotropic

liquid (TGA). DSC thermograms were qualitatively similar for

all compounds and in agreement with POM observations:

during the ﬁrst heating (and below 150 °C), weak and broad

exo- and endothermic signals were obtained, conﬁrming

complicated thermal history and phase coexistence behaviors.

Such complicated thermal behavior is commonly observed for

intricate supermolecular systems such as oligomers, polypedes,

and dendrimers.^30,44On subsequent heating−cooling cycles,

however (providing the decomposition temperatures were not

attained), DSC traces appeared mostly silent, and a glass

transition, corresponding to the freezing of the mesophase, was

systematically observed at low temperatures (Supporting

Information).

b

,

of a glassy LC phase and partially crystalline solid), consistent

with DSC and POM analyses, as evidenced by the presence of

several sharp, small- and large-angle reﬂections and broad

signals, particularly true for [Mn₁₂O₁₂(L_3,4,5‑CB)₁₆(H₂O)₄] and

[Mn₁₂O₁₂(L_3,5‑CB)₁₆(H₂O)₄]. Upon continuing heating (be-

tween ca. 80−120 °C), the abrupt complete disappearance of

one set of sharp reﬂections, those likely reminiscent of a

residual crystalline ordering, is observed, giving rise to diﬀuse

features, whereas another set of two sharp, small-angle

reﬂections still remains over the whole temperature range.

Once the samples were annealed for a few minutes in this 80−

120 °C temperature range to allow the complete melting of the

residual solid, clear diﬀractograms were obtained reversibly on

subsequent temperature cycles. To allow comparison, the same

thermal treatment was applied to the four samples, and in the

following, only data obtained from the second and third

temperatures cycles will be considered (no sign of degradation

being detected).

Temperature-dependent, small-angle X-ray diﬀraction experi-

ments (performed between 40 and 150−200 °C) permitted

one unequivocally to conﬁrm the formation and nature of the

LC mesophases for these systems, and to follow the evolution

of the phase parameters with temperature (Supporting

Information). Experiments carried out during the ﬁrst heating

revealed phases coexistence in the low-temperature ranges

consequently to low-kinetics transformations (i.e., coexistence

As expected, all dodecanuclear compounds self-organize into

lamellar smectic phases because the pending mesogens are

typical calamitic smectogens,²⁹and despite the diﬀerent ligands

substitution patterns, the smectic phases possess a similar

structure. As general features, all temperature-dependent

diﬀractograms are quite similar in all cases, showing sharp

reﬂections in the small-angle region, indicative of some long-

2712

dx.doi.org/10.1021/ja311190a | J. Am. Chem. Soc. 2013, 135, 2708−2722

Journal of the American Chemical Society

Article

range order supramolecular lattices, and two sets of intense and

diﬀuse signals with average maxima located at around 4.5−4.6

Å in wide-angle range, reﬂecting the liquid-like state of the

molten alkyl chains and the average packing of the disordered

diverging mesogenic subunits (h_ch+mes, Figure 3), and between

001 and 002 Miller indices (00l) (Table 2, Figure 3, Supporting

Information). The periodicities are of the same magnitude,

ranging between 42 and 47 Å (Table 1), and decrease linearly

as the number of pending cyanobiphenyl groups is changed

from 16 ([Mn₁₂O₁₂(L_4‑CB)₁₆(H₂O)₄]) to 48 ([Mn₁₂O₁₂-

(L_3,4,5‑CB)₁₆(H₂O)₄]) via 32 ([Mn₁₂O₁₂(L_3,x‑CB)₁₆(H₂O)₄], x =

4, 5) due to the overall molecular lateral expansion (Figures 3

and 4). This trend is a priori connected to speciﬁc packing

Figure 4. Plots of the relevant mesophase parameters as a function of

the substitution pattern of the Mn12 complexes.

constraints and dipolar interactions between the cyanobiphenyl

units (vide infra). For each compound, the lamellar periodicity

is also quasi-invariant with temperature (less than a few %),

suggesting little eﬀect of the temperature on the supramolecular

arrangements (Supporting Information).

The abnormal intensity proﬁle of the two reﬂections, with a

relatively weak (001) reﬂection with respect to (002), is a

strong indication of segregation occurring at the molecular level

between the cluster, the mesogens, and the aliphatic spacers,

respectively. Such an intensity distribution indeed arises from

the modulation of the electronic density within the lamellar

periodicity produced by the regular alternation of high- and

low-electronic density sublayers, and particularly here is

testimony (due to the enhanced 002 reﬂection) of an increase

of the electronic density in the central slab of the layer,

containing the strongly scattering polymetallic central cores.

Note that the variation of the relative intensities ratio observed

for the four complexes diﬀers slightly and obviously depends on

the respective thickness ratios of their diﬀerent sublayers (due

to the ligand substitution pattern), and on the sharpness of the

various interfaces. No obvious other change could be observed

Figure 3. Representative X-ray patterns of the three new clusters

recorded at 80 °C and of [Mn₁₂O₁₂(L_3,4,5‑CB)₁₆(H₂O)₄] at 120 °C

(top) and enlargement of the 1.5−6° 2θ-zone (bottom). Thin lines are

ﬁts of the broad diﬀusion line (h_cluster) with a Gaussian function (see

text).

ca. 19−24 Å, provisionally associated with a ribbon periodicity

and with the average correlation distance between ﬁrst

neighboring scattering polymetallic centers within the layers

(h_cluster, Figure 3 and Table 1); the mesophases can thus be

viewed as ﬂuid and rather disordered (no in-plane ordering of

the peripheral mesogenic units). The layer periodicities are

directly given by the position of the two sharp and equidistant

small-angle reﬂections in the spacing ratio 1:2, indexed with the

Table 2. Comparison of the Mean Distance between First Neighboring Clusters for Ideal Hexagonal and Square Symmetrical

a

Arrangements, XRD Values, and Molecular Dynamic Values

hexagonal symmetry

square symmetry

h

L in cluster

[Mn₁₂O₁₂(L)₁₆(H₂O)₄]

A_mol

h_min

⟨h_hex

⟩

h_max

h_min

⟨h_sq⟩

h_max

h_cluster

h_md(sq)

h_md(hex)

L_3,4,5‑CB

L_3,5‑CB

L_3,4‑CB

L_4‑CB

810

580

530

300

26.48

22.4

27.78

23.51

22.47

16.91

30.6

25.9

24.74

18.6

20.12

17.03

16.28

12.25

22.58

19.11

18.27

13.75

28.46

24.08

23.02

17.32

23.6 1.0

20.2 2.0

20.9 2.0

19.0 3.0

28.4

23.0

−

30.5

−

21.4

16.11

17.0

−

a

2

Hexagonal symmetry: A_mol= h_min·h_max, h_max/h_min= 2/√3, and ⟨h_hex⟩ = 3^5/4log 3 ×√A_mol/[π√2] = 0.9763√A_mol. Square symmetry: A_mol= (h_max

)

= 2 × (h_min)², h_max/h_min= √2, ⟨h_sq⟩ = 2√2 log(1 + √2) × √A_mol/[π] = 0.7935√A_mol; h_clusterwas measured by SAXS; h_mdwas obtained from

molecular dynamics considering a square lattice in all cases, and square and hexagonal lattices for [Mn₁₂O₁₂(L_3,4,5‑CB)₁₆(H₂O)₄]. ⟨h_hex⟩ and ⟨h_sq⟩ were

determined according to the integral equation: ⟨h/h_min⟩ = (θ₂− θ₁)⁻¹∫ dθ/cos θ, where θ₁and θ₂are the limits for the integration: θ₁= 0 and θ₂=

π/6 and π/4 for the hexagonal and square symmetry, respectively; (−) modeling not performed.

2713

dx.doi.org/10.1021/ja311190a | J. Am. Chem. Soc. 2013, 135, 2708−2722

Journal of the American Chemical Society

Article

in the X-ray patterns within the series of [Mn₁₂O₁₂-

(L_x,y,z‑CB)₁₆(H₂O)₄] clusters between 40 and 150/200 °C,

except for a strong temperature- and structure-dependency of

the relative intensity ratio of the two sharp reﬂections, I₀₀₁/I₀₀₂

,

which decreases on raising the temperature from approximately

ca. 4:1 to 1:1 or 1:2 (or even lower, because the 001 reﬂection

may be almost extinct); the phenomenon is completely

reversible with temperature on subsequent heat−cool cycles

(Figure 3).

Present for all samples, the small-angle broad diﬀusion

(h_cluster, Figure 3), localized at the base of the sharp 002

reﬂection, and whose value is slightly larger than the average

diameter of the naked cluster, corresponds indeed to a

periodicity and to the short-range correlation distance between

metallic centers within the lamellae. Conﬁrming this

assumption, the maximum of the diﬀusion, determined by

ﬁtting the experimental line shape with a Gaussian function,

also varies linearly with the terminal mesogenic groups number

(Figure 4), that is with the lateral molecular expansion.

Furthermore, in the investigated mesomorphic temperature

interval, this broad diﬀusion does not shift signiﬁcantly, and the

intensity of the signal remains more or less constant

(Supporting Information).⁴⁵The speciﬁc molecular structures

and these observations strongly support that this diﬀusion

actually belongs to a periodicity produced by the disruption of

the aliphatic sublayer by the hard-core cluster; in other words,

this signal arises from the modulation of the electronic density

in the sublayer consequent to the embedment of the highly

scattering clusters within the low-electron density aliphatic

continuum. Moreover, the relative broadness of the signal

indicates some limited and short-range in-plane correlations

between the cluster cores within the layer, such as, ﬁrst

neighbor’s type correlation, obviously not correlated with

adjacent layers (no 3D correlations).

Therefore, despite the structural variations within the

members of the series, all of the diﬀractograms recorded in

the mesophases temperature ranges look similar (Figure 3),

suggesting that the same mesophase structure is kept in the

whole series. As found for structurally related hybridized

molecular systems of this type,^{32,36−38,40,41}the mesophase can

thus be satisfactorily described by a smectic morphology,

consisting of a bilayer structure made of outer mesogens

sublayers with a short-range 2D in-plane arrangement of the

clusters embedded in the median aliphatic sublayer (vide infra).

As far as the local packing is concerned, subtle diﬀerences

depending on the compounds can be revealed, which are

evidently associated with the various ligands substitution

pattern topologies and consequently with the number/position

of appending mesogens connected to the clusters. These

modiﬁcations within the series can be appreciated by following

the evolution of the molecular areas, A_mol, deduced from the

ratio between the molecular volumes (V_mol), calculated with a

density of 1.00 0.05 g cm⁻³(Table 1), and the mean layer

thickness (d). As expected, the molecular area is found to

increase concomitantly with the number of end-mesogens,

evidencing the bulkiness of the periphery in comparison to the

cluster hard-cores (Figure 4). These values also prompt that the

fundamental lamellar periodicities therefore correspond to one

monomolecular bilayer, induced by the lateral two-dimensional

registry of the molecules in pseudocylindrical conformation,

that is, with one-half of the mesogenic pending arms distributed

on either side of the median molecular plane, all pointing on

average along the same direction (Figure 5). The mesogenic

Figure 5. Top: Schematic representation of the cylindrical shape (d ×

A_mol) of the hybrids and perspective view of the self-organization of

the Mn12 complexes into smectic layers. Bottom: Representations of

the two in-plane lattices (hexagonal and square symmetry,

respectively) for the 2D arrangements of the clusters within the

median sublayer and description of the relevant geometrical

parameters (the blue circles represent the time-averaged circular

projection of the cluster onto the lattice plane, the dotted circles for

the dual lattices, and the white parts the aliphatic continuum; the solid

and dotted light-blue lines represent the large and small (dual) lattices

of these arrangements, and the darker solid lines represent the ﬁrst

neighbors’ lattice); h_maxand h_min(black arrows) represent the

maximum and minimum distances between two nodes within the

lattice plane (hexagonal and/or square symmetry, i.e., the parameters

of the large and small (dual) lattices respectively); ⟨h_hex⟩ and ⟨h_sq⟩

(black dotted arrows) represent the average distances between lattice

nodes (see Table 2).

transverse cross-section area, a_mes(a_mes= 2A_mol/n_meswith n_mes

=

total number of cyanobiphenyl units per cluster, Table 1),

reveals a strong substitution-dependent compactness with

values ranging around 33 Å²for [Mn₁₂O₁₂(L_3,4,5‑CB)₁₆(H₂O)₄]

and [Mn₁₂O₁₂(L_3,4‑CB)₁₆(H₂O)₄] to about 37 Å²for [Mn₁₂O₁₂-

(L_3,5‑CB)₁₆(H₂O)₄] and [Mn₁₂O₁₂(L_4‑CB)₁₆(H₂O)₄] (Table 2).

First, all of these values are signiﬁcantly larger than the

theoretical cross section of an aliphatic chain (22.5 Å²),

expected in the ideal case of long-range lateral packing of

mesogens, parallel to the layer normal as in a classical SmA

phase. Such divergences usually originate from the combination

of partial mesogens’ interpenetration between adjacent layers

and/or from a substantial tilt of the rigid cores with respect to

the layer’s normal, leading to undulated interfaces. Second, and

speciﬁcally here, divergences also result from the non-

equivalence of the connecting sites around the rigid and

nondeformable cluster core, axial and equatorial positions,

which also contribute substantially to the variation of the

transverse area.

Clearly, the bulkier is the periphery, the more densely the

peripheral substituents are laterally close-packed in the

sublayers. Indeed, the polysubstituted systems, [Mn₁₂O₁₂-

2714

dx.doi.org/10.1021/ja311190a | J. Am. Chem. Soc. 2013, 135, 2708−2722

Journal of the American Chemical Society

Article

(L_3,4,5‑CB)₁₆(H₂O)₄] and [Mn₁₂O₁₂(L_3,4‑CB)₁₆(H₂O)₄], show a

stronger tendency than do the less crowded ones, [Mn₁₂O₁₂-

(L_4‑CB)₁₆(H₂O)₄] and [Mn₁₂O₁₂(L_3,5‑CB)₁₆(H₂O)₄] (because of

the nonoccupied meta or para positions, respectively) to depart

from the sublayers formed by the mesogens connected at the

axial ones, essentially by pulling those connected at the

equatorial positions, reducing the average molecular area per

mesogen relatively (Table 1). Interestingly, the two constitu-

tional isomers, [Mn₁₂O₁₂(L_3,5‑CB)₁₆(H₂O)₄] and [Mn₁₂O₁₂-

(L_3,4‑CB)₁₆(H₂O)₄], behave quite diﬀerently, and this behavior

seems to concord with the numerous possible conformations

adopted for the latter compound (consequent to its nonsym-

metrical substitution pattern), which can, in a ﬁrst approx-

imation, be time-averaged with [Mn₁₂O₁₂(L_3,4,5‑CB)₁₆(H₂O)₄]

compound. Thus, the evolution of the lamellar packing within

the series is fairly well understood by considering altogether the

number of pending mesogens (i.e., 48, 32, 16), the substitution

pattern (i.e., 3,4,5-, 3,4-, 3,5-, and 4-), and the two connecting

modes around the cluster (i.e., axial versus equatorial

positions), all inﬂuencing at various levels areas and packing

constraints.

As mentioned above, the segregated nature and cylindrical

conformation of the hybrids (Figure 5) imply also that the

bulky metallic hard-cores are somehow correlated in the

median plane of the smectic layer. Unfortunately, discussion

upon the exact lateral two-dimensional packing of central

clusters within the correlated single layers could only be

speculative, because only a diﬀuse signal (h_cluster, Figure 3)

could be detected in the powder patterns; the correlation is

only short-range, and extends to ﬁrst neighbors only. Generally,

and particularly here, the mean packing viewed by X-ray

diﬀraction actually emerges from the local scale molecular

packing, determined by interactions between neighboring

molecules, molecular conformational changes, and motions,

which is smoothed and averaged. Moreover, experiments on

aligned samples did not yield oriented patterns of the

mesophase, precluding the absolute assignment of the

symmetry and orientation of this pseudolattice. In 2D self-

assemblies, positional ordering is mostly hexagonal, but other

periodic patterns such as square, pentagonal tiling, etc., or even

binary tilling combinations, which increase further the

possibilities, are possible.^30e,iWhatever the symmetry consid-

ered for such a local arrangement, the geometry is determined

by the molecular area, which deﬁnes the lattice area. In the

present case, the molecules adopt a cylindrical shape, with a

time-averaged circular cross-section, and therefore essentially

the two likely possibilities of 2D patterning are square or

hexagonal (Figure 5). The partial deciphering of the local

packing and the estimation of ⟨h⟩, the mean average distance

between the two closest adjacent circular surfaces, which in our

case represents the average distance between two nearest

clusters, were achieved by using a straightforward geometrical

approach considering both hexagonal and square packing

comparable with the average distances between clusters

measured by SAXS (h_cluster) for each sample. In addition,

from these calculations, a slight trend toward the square lattice

is revealed (Table 2, h_cluster≈ ⟨h_sq⟩) for all compounds, a likely

consequence of the bulkiness of the periphery, except for the

less crowded [Mn₁₂O₁₂(L_4‑CB)₁₆(H₂O)₄] (Table 2, ⟨h_hex⟩ ≈

h_cluster). When the number of mesogens is increased, the

molecular area is dominated by the peripheral mesogens,

leading to more compact arrangements and hindering the

overall molecular rotation around the main molecular axis

(compare A_moland a_mesvalues in Table 1). In addition, due to

the large number of susbstituents, the appending gallates

irradiating from the core (equatorial positions) force the Mn12

clusters out of the median plane by pulling the mesogens

toward the outer mesogenic sublayers (for [Mn₁₂O₁₂-

(L_3,4,5‑CB)₁₆(H₂O)₄] and [Mn₁₂O₁₂(L_3,4‑CB)₁₆(H₂O)₄]). The

combined eﬀects of the hindered rotation, out of plane

orientation, and equatorial crowding contribute to the local

symmetry reduction and to the persistence of the overall 4-fold

symmetry, and consequently lead here preferentially to a

square-like arrangement. In the opposite situation, [Mn₁₂O₁₂-

(L_4‑CB)₁₆(H₂O)₄], the molecular area is less restricted by the

periphery; the terminal mesogens and the less crowded

equatorial benzoates have more freedom, and the cluster is

not pulled out of the plane: this peculiar arrangement likely

blurs the intrinsic S₄symmetry of the cluster by smoothing the

interface with the aliphatic continuum, and drives to both in-

plane arrangements equally. For [Mn₁₂O₁₂(L_3,5‑CB)₁₆(H₂O)₄],

the intermediate situation occurs (large a_mes, no tilted core, and

large number of substituents).

Validation of this supramolecular arrangement was attempted

by molecular modeling and molecular dynamic calculations.

The simulation methods used here provide a snapshot of the

conformations and arrangements of few-molecules models in a

minimized energy status. The detailed procedure of the

simulation⁴⁶consisted of placing two molecules (two super-

imposed subsets of molecules), formerly optimized (e.g., with a

more or less cylindrical shape), in an orthorhombic box and

setting a constraint onto the molecular area A_moldeduced from

X-ray data and volumetric calculations. The z parameter of the

orthorhombic cell was set to 200 Å (much larger than twice the

measured periodicity, Table 1) so as not to constrain the

thickness artiﬁcially and to allow suﬃciently large molecular

motions for an eﬃcient and reproducible optimization of the

conformations and of the supramolecular arrangements. Each

manganese atom of a given cluster was restricted to its position

relative to the others within the metallic framework, according

to the X-ray crystalline structure of the trimethoxy gallate

homologous cluster,³²whereas all other atoms were allowed to

move freely, as were the two molecules in the constrained cell.

Modeling results previously obtained for [Mn₁₂O₁₂-

(L_3,4,5‑CB)₁₆(H₂O)₄] supported this view of the self-assembly

of highly segregated molecules into bilayer smectic phases, in

good accordance with the diﬀraction parameters, and

particularly they highlighted the good convergence of the

lattice dimension toward the lamellar periodicity. Molecular

dynamics further highlighted that all of the clusters cores were

tilted out of the plane with respect to the median sublayer of

the smectic phase with no tilt correlation within and between

layers (random orientation), conﬁrming the pulling eﬀect of the

peripheral mesogens. A square lattice was also preferred at that

time with respect to the hexagonal lattice (only these two

modes, corollary to the molecular cross-section areas, A_mol

.

Because the system is liquid-crystalline and thus ﬂuid, the ⟨h⟩

distance (⟨h_hex⟩ or ⟨h_sq⟩) ﬂuctuates logically between two limit

values, h_minand h_max(h_maxis the largest distance and h_minis the

shortest distance between two neighbors, for each symmetry

considered, Figure 5). In fact, for each symmetry, h_maxand h_min

both correspond to the parameters of both the large and the

small (dual) 2D lattices, respectively. The minimum and

maximum (h_minand h_max) and mean (⟨h_sq⟩,⟨h_hex⟩) distances

between ﬁrst neighbors are in general good agreement and

2715

dx.doi.org/10.1021/ja311190a | J. Am. Chem. Soc. 2013, 135, 2708−2722

Journal of the American Chemical Society

Article

Figure 6. Side-views of the lamellar packing of the clusters (top) and top-views of the in-layer lateral arrangement of the clusters in square-like lattice

(bottom): [Mn₁₂O₁₂(L_3,5‑CB)₁₆(H₂O)₄] (left) and [Mn₁₂O₁₂(L_4‑CB)₁₆(H₂O)₄] (right).

lattices were considered, vide supra), also in agreement with the

overall intrinsic S₄molecular symmetry of Mn12 complexes.³²

New such simulations experiments were therefore performed

on the symmetrical clusters [Mn₁₂O₁₂(L_3,5‑CB)₁₆(H₂O)₄] and

[Mn₁₂O₁₂(L_4‑CB)₁₆(H₂O)₄] to conﬁrm the bilayer nature of the

structure, and eventually the nature of the lateral packing; the

nonsymmetrical nature of the other cluster precluded such

experiments. Some of the more relevant results of these

calculations are displayed in Figure 6. After important

oscillations in the beginning of the simulation, the amplitude

of the layers ﬂuctuations reaches a stable regime with

thicknesses converging to the lamellar values measured by

SAXS (Supporting Information), independent of the lattice

considered. Molecular dynamics was able to discriminate

between square and hexagonal lattices [Mn₁₂O₁₂-

(L_3,5‑CB)₁₆(H₂O)₄] (Figure 6) for which a good convergence

was observed (h_cluster≈ h_md(sq) = h_max(sq) Table 2) by setting a

square lattice, but unable to do so for [Mn₁₂O₁₂-

(L_4‑CB)₁₆(H₂O)₄] (h_md(sq) = ⟨h_hex⟩ and h_cluster= h_max(hex),

Table 2), for which the registry in the third direction was

necessary to reconcile the square arrangement with the

molecular area. As assumed above, the reality probably is an

average representation of both (expected by the broadness of

the diﬀuse X-ray signals), presumably resulting from slight

shifts in position (and orientation) between neighboring

clusters, further blurred out by the dynamics of the liquid

crystal phase (ﬂuidity of the mesophase). Finally, for

[Mn₁₂O₁₂(L_3,5‑CB)₁₆(H₂O)₄], molecular dynamics revealed

that the clusters are not inclined out of the median plane,

unlike for [Mn₁₂O₁₂(L_3,4,5‑CB)₁₆(H₂O)₄], but remain rather ﬂat

in the plane as for the less crowed system, [Mn₁₂O₁₂-

(L_4‑CB)₁₆(H₂O)₄], in agreement with the larger (and similar)

values of a_mes(Table 1).

To conclude on the mesomorphic properties, all of the

clusters self-organize at room temperature into smectic bilayers,

with the clusters fully embedded in the central aliphatic

sublayers, forming a weakly correlated 2D lattice. It was shown

that the substitution pattern can be used as a means to control

the packing density and stimulate the type of symmetry of the

local 2D clusters arrangement (a_mes, tilting, equatorial

crowding). Of particular interest, in the case of the densely

packed systems, is the tendency of the equatorial mesogenic

groups to reach the mesogenic sublayers, forcing the clusters

out of plane and probably inducing important local structural

distortions of the clusters.

Magnetic Properties of the Dodecanuclear Clusters

[Mn₁₂O₁₂(L_x,y,z‑CB)₁₆(H₂O)₄]. The signature of SMM com-

pounds is generally associated with the presence of a frequency-

dependent out-of-phase ac susceptibility signal (χ′′) that can be

measured when the frequency of the ac ﬁeld becomes close to

the relaxation frequency of the molecule magnetic moment.⁴⁷

The ac magnetic susceptibility is also used to determine the

spin of the ground state (assuming that only the ground state is

populated at this temperature) as well as the height of the

eﬀective energy barrier (U_eff), which prevents the thermal

relaxation of the magnetization. Indeed, the relaxation time of

2716

dx.doi.org/10.1021/ja311190a | J. Am. Chem. Soc. 2013, 135, 2708−2722

Journal of the American Chemical Society

Article

Table 3. Magnetic Properties for [Mn₁₂O₁₂(L_x,y,z‑CB)₁₆(H₂O)₄]

L in cluster

[Mn₁₂O₁₂(L)₁₆(H₂O)₄]

a

b

c

d

e

f

relaxation modes

T_B(K)

μ₀H_C(T)

S, g

U_eff(K)

τ₀(s)

JT isomeric ratio (SR/FR)

L_3,4,5‑CB

L_3,5‑CB

L_4‑CB

FR and SR

2.4

2.6

2.5

0.320

0.600

0.575

0.130

9, 1.91

9, 1.99

10, 1.88

9, 1.87

55

60

59

1.4 × 10⁻⁸

8.4 × 10⁻⁹

6.2 × 10⁻⁹

4.4 × 10⁻⁹

65/35

60/40

55/45

L_3,4‑CB

35/65

a

b

SR: slow relaxing phase. FR: fast relaxing phase. The FR phases occur at too low temperature and/or are too weak to be fully characterized. T_B:

c

blocking temperature (arbitrary deﬁned as corresponding to a relaxation time of 100 s for the SR species). μ₀H_C: coercive ﬁeld measured at T = 1.8

K. S, g: spin and Lande factor of the ground state. U_eff: potential energy barrier. τ₀: pre-exponential factor.

d

e

f

Figure 7. Temperature dependence of χ_acT (left column) and χ″ (right column) for, from top to bottom, [Mn₁₂O₁₂(L_3,4,5‑CB)₁₆(H₂O)₄],

[Mn₁₂O₁₂(L_3,5‑CB)₁₆(H₂O)₄], [Mn₁₂O₁₂(L_3,4‑CB)₁₆(H₂O)₄], and [Mn₁₂O₁₂(L_4‑CB)₁₆(H₂O)₄]. χ_ac= (χ′²+ χ″²)^0.5, with χ′ and χ″ being, respectively, the

real and imaginary components of the molar magnetic susceptibility (χ = χ′ + iχ″) at zero dc ﬁeld with a ﬁeld of 3.5 Oe oscillating at 1000 (black),

500 (red), 100 (cyan), 50 (green), 10 (magenta), and 1 Hz (dark blue).

SMM magnetization usually follows a thermally activated

process that can be described by the Arrhenius law (τ = τ₀

exp(−U_eff/k_BT), where τ is the relaxation rate, k_Bis the

Boltzmann constant, and τ₀is a pre-exponential factor), making

2717

dx.doi.org/10.1021/ja311190a | J. Am. Chem. Soc. 2013, 135, 2708−2722

Journal of the American Chemical Society

Article

thus possible the determination of the blocking temperature

(T_B, arbitrarily deﬁned as corresponding to a relaxation time of

100 s). For [Mn₁₂(OAc)₁₆], the parameters used to ﬁt the

Arrhenius type dependence of the relaxation time of the

magnetization, τ, are: τ₀= 2.1 × 10⁻⁷s and U_eff/k_B= 61 K.¹⁵

This means that the relaxation time of the magnetization of

[Mn₁₂(OAc)₁₆] is of the order of months at 2 K. The fact that

the low-temperature slow relaxation of the magnetization of

SMM occurs at the molecular level and does not result from

collective interactions has been experimentally established, in

particular with low-temperature calorimetry (no speciﬁc heat

anomaly) and magnetometry on diluted samples. As none of

the clusters involved in this study could be obtained as large

single crystals, essentially because of the constitutive ﬂexible

and mesomorphous organic coating, the study of the magnetic

properties was performed on powder samples instead, still

admitting the hypothesis of the above relaxation pathway. The

same thermal treatment as described above (SAXS) was applied

here before carrying magnetic measurements to bring the

samples in the same homogeneous state, and thus to eventually

allow precise structural comparisons. Note that prior magnetic

measurements performed without thermal treatment gave very

similar results, and the values found in this set of compounds

are in good agreement with what is usually found for Mn12

clusters, suggesting that (i) the integrity of the clusters is

preserved during thermal annealing, and (ii) the coexisting

smectic mesophase and partially crystallized solid (vide supra)

likely possess a similar type of supramolecular lamellar

(pre)organization. All of these parameters gained from χ″

and χ_acT (χ_ac= (χ′²+ χ″²)^1/2) versus T and magnetization

hysteresis cycles measurements are summarized in Table 3 (and

Supporting Information). In Figure 7, plots of χ_acT product and

χ″ versus T at various oscillating frequencies (from 1 kHz to 1

Hz) are displayed. The frequency-dependent decrease in the

χ_acT versus T plot on lowering the temperature, observed in all

four Mn12 derivatives, indicates that the relaxation of the

magnetization rate becomes close to the ac ﬁeld frequency.

Accordingly, a χ″ signal appears at the corresponding

temperature, indicating SMM behavior for all of the Mn12-

based materials.

For the three novel compounds, the χ_acT values above 8 K

are equal and constant at all frequencies. Assuming that only

the ground state is populated at this temperature, the value of

the χ_acT product can be used to determine the spin of the

ground state of the clusters. The values of the χ_acT products

measured at the plateau were 44.5, 39.5, and 48.8 emu K mol⁻¹,

leading to S = 9 and g = 1.99, S = 9 and g = 1.87, and S = 10

and g = 1.88 for [Mn₁₂O₁₂(L_3,5‑CB)₁₆(H₂O)₄], [Mn₁₂O₁₂-

(L_3,4‑CB)₁₆(H₂O)₄], and [Mn₁₂O₁₂(L_4‑CB)₁₆(H₂O)₄], respec-

tively, in fair accordance with those determined for other

Mn12-based complexes: even though most of them possess a S

= 10 spin ground state, systems with S = 9 are not rare and have

been measured in many cases (vide supra).⁴⁸Recall that for

[Mn₁₂O₁₂(L_3,4,5‑CB)₁₆(H₂O)₄], the value of χ_acT at the plateau

was around 41 emu K mol⁻¹, corresponding to a ground state

of S = 9 and g = 1.91.³²The onset of a second plateau is

observed at lower temperatures (χ_acT ≤ 20 emu K mol⁻¹,

Figure 7).

cesses, the ﬁrst one in the region 4−8 K, corresponding to a

slow relaxation mode (SR), and the other one in the region 1−

3 K, corresponding to a fast relaxation mode (FR) (Figure 7).

The occurrence of two peaks in the out-of-phase susceptibility

is not an unusual feature for Mn12 derivatives.⁴⁹A well-

accepted explanation for the presence of two relaxation modes

is due to subtle structural diﬀerences in the molecular

structures, not damaging the clusters. One structural diﬀerence

may arise from the position of the four water molecules

coordinated to type II Mn^IIIions,^11,48,50,51but the most

signiﬁcant diﬀerence at the origin of these two relaxation modes

is attributed to JT isomerism, due to the splitting of the

degenerate d orbitals of the Mn^IIIions (d⁴) in octahedral ligand

ﬁeld (O_h), which elongation leads to a tetragonal geometry

(D_4h).¹⁹At this stage, because no crystallographic structure is

available, we propose the hypothesis that the presence of these

two peaks is principally caused by the coexistence of JT isomers

within each sample. The ratio of each isomer can be

approximately determined by the diﬀerence of the two plateaus

present in χ_acT versus T plots. The SR/FR ratios measured are

ca. 65/35, 60/40, 55/45, and 35/65 for Mn12 systems with

ligands substitution in the decreasing sequence L_3,4,5‑CB, L_3,5‑CB

L_4‑CB, and L_3,4‑CB, respectively.

,

The dynamic magnetic behavior of the complexes [Mn₁₂O₁₂-

(L_x,y,z‑CB)₁₆(H₂O)₄] seems to be somehow inﬂuenced by the

overall molecular structure, that is, more precisely, by the

peripheral ligand substitution patterns. A careful examination of

the Mn^IIIcoordination spheres of Mn12 structures indicates

that whatever the positional isomer, the number of

coordination sites substituted by carboxylates oxygen’s is

constant (i.e., 16 equatorial and 12 axial positions for the 8

Mn^IIIions). When the acetates of the parent [Mn₁₂(OAc)₁₆]

cluster are replaced by the benzoate ligands, the crystal ﬁeld

around Mn^IIIions is substantially modiﬁed. This is crucial for

the d⁴Mn^IIIions, showing a strong tendency toward JT

distortion. Therefore, the substitution of acetates by ligands

providing higher crystal ﬁeld will globally stabilize the JT

elongation along the main symmetry axis of the clusters. This is

the case with these ligands, as it appears that the stabilization of

the SR phase with respect to the FR depends also on the

molecular symmetry and follows the trend HL_3,4,5‑CB≥ HL_3,5‑CB

≥ HL_4‑CB> HL_3,4‑CB. Indeed, while for the three symmetrical

systems, the SR magnetic phase remains predominant and the

FR almost nonexistent, the relative abundance of the two

phases is inverted in the nonsymmetrical [Mn₁₂O₁₂-

(L_3,4‑CB)₁₆(H₂O)₄] complex. Interestingly too, in the sym-

metrical systems, that is, with ligands L_3,4,5‑CB, L_3,5‑CB, L_4‑CB, this

trend also is observed conjunctly with the decrease of the

number of ether functions attached to the benzoate moieties

coordinated to the Mn^IIIions. As ether functions act as

electron-donating groups, the electronic density of the aromatic

ring is altered, and consequently the crystal ﬁeld of the

benzoate anchors is modiﬁed accordingly (vide supra,

Supporting Information).

For the four complexes, the magnetization relaxation rate τ

follows the Arrhenius equation (Figure 8). Only the parameters

corresponding to the high-temperature peak (slow relaxing

phase, SR) could be extracted with suﬃcient precision for each

sample from the peak maxima of the χ″ versus T plots (Table

3). All fall within the expected ranges for Mn12 complexes: τ₀

lies in the interval of 5−10 ns; the eﬀective energy barrier is

comprised between 55 and 60 K.

As evidenced by the presence of two plateaus in the χ_acT

versus T products and two peaks in the χ″ versus T plots, two

frequency-dependent out-of-phase ac susceptibility signals were

also observed for the four compounds [Mn₁₂O₁₂-

(L_x,y,z‑CB)₁₆(H₂O)₄], suggesting two diﬀerent relaxation pro-

2718

dx.doi.org/10.1021/ja311190a | J. Am. Chem. Soc. 2013, 135, 2708−2722

Journal of the American Chemical Society

Article

highest μ₀H_Cvalue is found for the symmetrical systems (those

having the highest SR/FR ratio), and second this value

increases along the sequence L_3,4‑CB< L_3,4,5‑CB< L_3,5‑CB

≈

L

_4‑CB, that with ligand L_3,4‑CBgiving rise to multiple regio-

isomers likely to contribute even more to the narrowing of the

hysteresis loop at zero ﬁeld.

DISCUSSION

■

All four Mn12 clusters have retained their original SMM

magnetic behavior upon hybridization by structuring organic

ligands, and the characteristic parameters are those expected for

this archetypical mixed-valent cluster.¹²We noted however two

subtle diﬀerences (recall that all measurements were performed

under identical experimental conditions in all cases): (i) about

the relative proportion of both SR and FR species, which

appear to be correlated with the structure of the ligand (e.g., the

symmetry of the substitution pattern), that is the overall

symmetry of the complex and number of geometrical isomers,

and (ii) concerning the large variations of the coercive ﬁeld

values (0.1−0.6 T), again correlated to the substitution pattern

of the ligands, which aﬀects the proximity of the clusters within

the median plane of the smectic layers, on the one hand, and

their relative out of plane orientation with respect to the layer

normal, on the other hand, consequently to the pulling eﬀect.

At this stage, it appeared thus pertinent to try ﬁnding trends

or correlations between the magnetism of these LC hybridized

Mn12 clusters and their arrangement within the mesophases.

Obviously, this task is tricky, because events occurring at

diﬀerent temperatures (low-temperature magnetic properties

versus room-temperature supramolecular organization) are

compared. However, consequent to thermal annealing and

rapid cooling, all mesomorphous clusters were analyzed under

the same experimental conditions, once into their glassy

smectic phase, which retains essentially the ordering and

packing integrity of the mesophase. After the low-temperature

magnetic measurements, the samples were removed from the

sample holders, and analyzed by DSC and SAXS (from −50 to

ca. 100 °C), revealing unmodiﬁed materials, and most

importantly with the same thermal behaviors (identical DSC

traces and glass transition temperatures) and supramolecular

organizations (smectic mesophase) as before the experiments.

Primarily, the relative proportion between SR and FR species

seems to be dependent on the substitution pattern of the

ligands, as all of the symmetrical complexes exhibit a higher

proportion of SR species than FR ones, while this ratio is

inverted in the case of the nonsymmetrical complex (with 3,4-

substitution). It is not surprising that the larger number of

geometrical isomers generated by this complex would likely

perturb the SR/FR ratio (vide supra). Secondary, another

striking feature is the variation of the coercive ﬁeld along the

series, considering that the same experimental conditions were

systematically reproduced for the measurements. From the

above analyses, the two compounds that have the lowest

coercivity are those whose cores are tilted out of the median

plane of the smectic layers as deduced by SAXS and are the

more compact systems (Figure 10), and it is not outrageously

speculative to emphasize that the pulling or pressure eﬀects of

the peripheral mesogens (to reach the mesogenic outer

sublayer) distort the cluster and aﬀect its magnetic properties.

Although the eﬀect of the true nature of the 2D lattice

symmetry (local in-plane arrangement of the clusters), which

could not be unequivocally assigned (hexagonal versus square),

on the magnetic properties could not be discriminated, an

Figure 8. Plots of the natural logarithm of relaxation rate, ln(1/τ),

versus inverse temperature for [Mn₁₂O₁₂(L_3,4,5‑CB)₁₆(H₂O)₄] (black),

[Mn₁₂O₁₂(L_3,5‑CB)₁₆(H₂O)₄] (red), [Mn₁₂O₁₂(L_3,4‑CB)₁₆(H₂O)₄]

(green), and [Mn₁₂O₁₂(L_4‑CB)₁₆(H₂O)₄] (cyan) using the χ″ versus

T data for the SR species of Figure 7 and Table 3. Full lines

correspond to the best ﬁts obtained from an Arrhenius law (see text).

Finally, the shape of the magnetization versus ﬁeld curves

recorded at 1.8 K at about 700 G min⁻¹for the four hybridized

compounds is typical for randomly oriented polycrystalline

Mn12 species (Figure 9).²³The magnetization curves exhibit

Figure 9. Magnetization versus applied ﬁeld for [Mn₁₂O₁₂-

(L_3,4‑CB)₁₆(H₂O)₄] (green), [Mn₁₂O₁₂(L_3,4,5‑CB)₁₆(H₂O)₄] (black),

[Mn₁₂O₁₂(L_3,5‑CB)₁₆(H₂O)₄] (red), and [Mn₁₂O₁₂(L_4‑CB)₁₆(H₂O)₄]

(cyan) at 1.8 K (the right ﬁgure is a zoom of the central part of the

ﬁgure on the left side).

an opened hysteresis loop characteristic of SMM behavior with

a coercive ﬁeld and a remnant magnetization (the absence of a

saturation of the magnetization at high ﬁelds is due to the

polycrystalline nature of the materials). The coercive ﬁeld

values at 1.8 K are in the range 0.13−0.6 T, in accordance with

the relative proportion of the SR species versus the FR ones:

the largest values (0.6−0.32 T) are indeed found for the

derivatives having the largest SR species content (with respect

to the FR one), whereas the smallest value (0.13 T) is obtained

with the compound derived from L_3,4‑CBligand in which the JT

distorted isomer (FR) is predominant. Moreover, one can

discern that the variation of the measured values of the coercive

ﬁeld (and to some extend the remnant magnetization) for each

sample seems also to be, at ﬁrst sight, correlated to the ligand

substitution pattern that is the number and position

(symmetry) of mesogenic end-groups (Table 1): ﬁrst the

2719

dx.doi.org/10.1021/ja311190a | J. Am. Chem. Soc. 2013, 135, 2708−2722

Journal of the American Chemical Society

Article

essential for realistic technological applications. Last, still in

view of the aforementioned development of a molecular

magnetic memory, one may imagine its 3D implementation

within such mesophases, taking proﬁt of the original photo-

sensitivity of Mn12 derivatives.³⁵In addition to the density gain

resulting from using bit elements much smaller than the current

ones, one would also hugely beneﬁt from going from 2D to 3D.

ASSOCIATED CONTENT

* Supporting Information

Experimental section including techniques, syntheses of ligands

■

S

1

and complexes, IR (Figures S1−S3) and H NMR (Figures

S4−S6) spectroscopies, DSC (Figures S7−S10, S22−S25) and

TGA (Figure S21) curves, POM images (Figures S11−S20),

XRD patterns (Figures S26−S28), molecular dynamics

(Figures S29,S30), Jahn−Teller distortion diagrams in

octahedral complexes (Figure S31), and tables (S1−S3). This

material is available free of charge via the Internet at http://

pubs.acs.org.

Figure 10. Evolution of the transverse molecular area, a_mes, and

coercive ﬁeld, μ₀H_C, as a function of the ligand-substitution patterns.

interesting observation however is that the further apart are the

clusters, the smaller is the coercive ﬁeld. It has been previously

shown that intermolecular magnetic interactions were of

outmost importance in the behavior of SMM, especially in

the quantum tunneling of magnetization (QTM), which is

indeed not purely molecular in origin.⁵²The present work

shows that it might be possible to ﬁnely tune the behavior of

SMM, by allowing precise control of the intercluster distances

(of the order of 20 Å) and their respective orientation.

AUTHOR INFORMATION

Corresponding Author

bdonnio@ipcms.u-strasbg.fr

■

Present Address

^†Department of Inorganic Chemistry, University of Geneva, 30

quai E. Ansermet, CH-1211 Geneva 4, Switzerland.

Notes

CONCLUSION

■

The authors declare no competing ﬁnancial interest.

In this study, we have reported on various SMM LC-hybridized

clusters able to self-organize into a room-temperature smectic

phase, which can be described by a bilayer structure made of

outer mesogens sublayers with a short-range 2D in-plane

arrangement of the clusters embedded in the median aliphatic

sublayer. This study revealed that this molecular design, namely

the grafting of mesogens around the Mn12 hard core, permitted

subtle structural modiﬁcations without drastically transforming

the mesophase organization (periodicity and compactness

variations depending on the substitution pattern, enhanced

thermal stability) and produced interesting new eﬀects on the

dynamic magnetic behavior (relative proportion of SR/FR

species and coercive ﬁelds values), which may be considered for

the design of future single molecule magnetic materials. All four

Mn12 clusters have indeed retained their original SMM

magnetic behavior upon hybridization by the structuring

organic ligands, and the characteristic parameters expected for

the this archetypical mixed-valent cluster could be ﬁnely tuned

by the speciﬁc molecular design and supramolecular organ-

ization. To further develop these fundamental aspects and to

optimize the various magnetic and mesomorphic properties,

one will obviously have to design new LC Mn12, or to extend

and adapt this strategy to other types of SMM clusters. For

instance, taking advantage on the diﬀerent reactivity of the

equatorial and axial substituents,⁵³potentially very interesting

new molecular heterogeneous Mn12 hybrids bearing mixed

(LC) ligands, with a high control of their distribution on both

positions, showing uncommon supramolecular arrangements

could be imagined. Concomitantly, many other fascinating and

versatile molecular structures derived from mixed d−f

complexes to high-nuclearity families of homo- and hetero-

metallic based-materials could be sought and designed to form

LC mesophases, appropriate mesogenic ligands being easily

available, and emerge with unprecedented features, for example,

higher temperature functioning being the most desirable one,

ACKNOWLEDGMENTS

This work was supported by the Swiss National Science

Foundation (E.T.), the CNRS (J.-L.G., G.R., B.D.), and the

Universite de Strasbourg. We thank Dr. Cyril Bourgogne (MD)

́

and Mr. Didier Burger (TGA).

■

REFERENCES

■

(1) Winpenny, R., Ed. Single-Molecule Magnets and Related

Phenomena. Structure and Bonding; Springer: Berlin, 2006; Vol. 122.

(2) Brooker, S.; Kitchen, J. A. Dalton Trans. 2009, 7331−7340.

(3) (a) Leuenberger, M. N.; Loss, D. Nature 2001, 410, 789−793.

(b) Cavallini, M.; Biscarini, F.; Gomez-Segura, J.; Ruiz, D.; Veciana, J.

Nano Lett. 2003, 3, 1527−1530. (c) Wernsdorfer, W. Nat. Mater.

2007, 6, 174−176. (d) Bogani, L.; Wernsdorfer, W. Nat. Mater. 2008,

7, 179−186. (e) Stepanenko, D.; Trif, M.; Loss, D. Inorg. Chim. Acta

2008, 361, 3740−3745. (f) Winpenny, R. E. P. Angew. Chem., Int. Ed.

2008, 47, 7992−7994. (g) Wernsdorfer, W. Int. J. Nanotechnol. 2010,

7, 497−522.

(4) Vincent, R.; Klyatskaya, S.; Ruben, M.; Wernsdorfer, W.;

Balestro, F. Nature 2012, 488, 357−360.

́

(5) (a) Gomez-Segura, J.; Veciana, J.; Ruiz-Molina, D. Chem.

Commun. 2007, 3699−3707. (b) Cavallini, M.; Facchini, M.; Albonetti,

C.; Biscarini, F. Phys. Chem. Chem. Phys. 2008, 10, 784−793.

(6) Rogez, G.; Donnio, B.; Terazzi, E.; Gallani, J.-L.; Kappler, J.-P.;

Bucher, J.-P.; Drillon, M. Adv. Mater. 2009, 21, 4323−4333.

(7) (a) Gatteschi, D.; Sessoli, R.; Cornia, A. Chem. Commun. 2000,

725−732. (b) Winpenny, R. E. P. J. Chem. Soc., Dalton Trans. 2002, 1−

10. (c) Christou, G. Polyhedron 2005, 24, 2065−2075. (d) Brechin, E.

K. Chem. Commun. 2005, 5145−5153. (e) Aromi, G.; Brechin, E. K.

Struct. Bonding (Berlin, Ger.) 2006, 122, 1−67. (f) Collison, D.;

McInnes, E. J. L.; Brechin, E. K. Eur. J. Inorg. Chem. 2006, 2725−2733.

(g) Escuer, A.; Aromi, G. Eur. J. Inorg. Chem. 2006, 4721−4736.

(h) Evangelisti, M.; Luis, F.; De Jongh, L. J.; Affronte, M. J. Mater.

Chem. 2006, 16, 2534−2549. (i) Bircher, R.; Chaboussant, G.; Dobe,

C.; Gudel, H. U.; Ochsenbein, S. T.; Sieber, A.; Waldmann, O. Adv.

̈

́

Funct. Mater. 2006, 16, 209−220. (j) Roubeau, O.; Clerac, R. Eur. J.

2720

dx.doi.org/10.1021/ja311190a | J. Am. Chem. Soc. 2013, 135, 2708−2722

Journal of the American Chemical Society

Article

Inorg. Chem. 2008, 4325−4342. (k) Tasiopoulos, A. J.; Perlepes, S. P.

Dalton Trans. 2008, 5537−5555. (l) Atanasov, M.; Comba, P.;

Hausberg, S.; Martin, B. Coord. Chem. Rev. 2009, 253, 2306−2314.

(24) Boskovic, C.; Brechin, E. K.; Streib, W. E.; Folting, K.; Bollinger,

J. C.; Hendrickson, D. N.; Christou, G. J. Am. Chem. Soc. 2002, 124,

3725−3736.

(25) Chakov, N. E.; Lee, S.-C.; Harter, A. G.; Kuhns, P. L.; Reyes, A.

P.; Hill, S. O.; Dalal, N. S.; Wernsdorfer, W.; Abboud, K. A.; Christou,

G. J. Am. Chem. Soc. 2006, 128, 6975−6989.

(m) Kortz, U.; Muller, A. J.; van Slageren, J.; Schnack, N. S.; Dalal, M.

̈

D. Coord. Chem. Rev. 2009, 253, 2315−2327. (n) Murrie, M. Chem.

Soc. Rev. 2010, 39, 1986−1995. (o) Kostakis, G. E.; Akoa, A.; Powell,

A. K. Chem. Soc. Rev. 2010, 39, 2238−2271. (p) Sessoli, R.; Powell, A.

K. Coord. Chem. Rev. 2010, 253, 2328−2341. (q) Kostakis, G. E.;

Hewitt, I. J.; Ako, A. M.; Mereacre, V.; Powell, A. K. Philos. Trans. R.

Soc., A 2010, 368, 1509−1536. (r) Wang, X.-Y.; Avendano, C.;

Dunbar, K. R. Chem. Soc. Rev. 2011, 40, 3213−3238. (s) Inglis, R.;

Milios, C. J.; Jones, L. F.; Piligkosd, S.; Brechin, E. K. Chem. Commun.

2012, 48, 181−190.

(26) Milios, C. J.; Inglis, R.; Bagai, R.; Wernsdorfer, W.; Collins, A.;

Moggach, S.; Parsons, S.; Perlepes, S. P.; Christou, G.; Brechin, E. K.

Chem. Commun. 2007, 3476−3478.

(27) Cirera, J.; Ruiz, E.; Alvarez, S.; Neese, F.; Kortus, J. Chem.-Eur. J.

2009, 15, 4078−4087.

(28) Tuan, N. A.; Katayama, S.; Chi, D. H. Phys. Chem. Chem. Phys.

2009, 11, 717−729.

(29) Handbook of Liquid Crystals; Demus, D., Goodby, J. W., Gray, G.

W., Spiess, H.-W., Vill, V., Eds.; Wiley-VCH: Weinheim, 1998.

(30) (a) Tschierske, C. Annu. Rep. Prog. Chem., Sect. C: Phys. Chem.

(8) Murugesu, M.; Habrych, M.; Wernsdorfer, W.; Abboud, K. A.;

Christou, G. J. Am. Chem. Soc. 2004, 126, 4766−4767.

́

(9) Ako, A.; M. Hewitt, I. J.; Mereacre, V.; Clerac, R.; Wernsdorfer,

2001, 97, 191−267. (b) Binnemans, K.; Gorller-Walrand, C. Chem.

̈

W.; Anson, C. E.; Powell, A. K. Angew. Chem., Int. Ed. 2006, 45, 4926−

4929.

Rev. 2002, 102, 2303−2345. (c) Kato, T.; Mizoshita, N.; Kishimoto, K.

Angew. Chem., Int. Ed. 2006, 45, 38−68. (d) Donnio, B.; Buathong, S.;

Bury, I.; Guillon, D. Chem. Soc. Rev. 2007, 36, 1495−1513.

(e) Tschierske, C. Chem. Soc. Rev. 2007, 36, 1930−1970. (f) Goodby,

(10) Lis, L. Acta Crystallogr., Sect. B 1980, 36, 2042−2046.

(11) (a) Sessoli, R.; Gatteschi, D.; Caneschi, A.; Novak, M. A. Nature

1993, 365, 141−143. (b) Sessoli, R.; Tsai, H.-L.; Schake, A. R.; Wang,

S.; Vincent, J. B.; Folting, K.; Gatteschi, D.; Christou, G.; Hendrickson,

D. N. J. Am. Chem. Soc. 1993, 115, 1804−1816.

J. W.; Saez, I. M.; Cowling, S. J.; Gortz, V.; Draper, M.; Hall, A. W.;

̈

Sia, S.; Cosquer, G.; Lee, S.-E.; Raynes, E. P. Angew. Chem., Int. Ed.

2008, 47, 2754−2787. (g) Bruce, D. W. Struct. Bonding (Berlin) 2008,

126, 161−180. (h) Bisoyi, H. K.; Kumar, S. Chem. Soc. Rev. 2011, 40,

306−319. (i) Tschierske, C. Top. Curr. Chem. 2012, 318, 1−108.

(31) (a) Kumar, S. Chem. Soc. Rev. 2006, 35, 83−109. (b) Laschat, S.;

(12) (a) Mertes, K. M.; Suzuki, Y.; Sarachik, M. P.; Myasoedov, Y.;

Shtrikman, H.; Zeldov, E.; Rumberger, E. M.; Hendrickson, D. N.;

Christou, G. Solid State Commun. 2003, 127, 131−139. (b) Bagai, R.;

Christou, G. Chem. Soc. Rev. 2009, 38, 1011−1026.

Baro, A.; Steinke, N.; Giesselmann, F.; Hagele, C.; Scalia, G.; Judele,

̈

(13) The S = 10 ground state is further split into 11 M_Ssublevels

(i.e., 2S + 1) in zero ﬁeld, and the M_S= 10 levels are the lowest in

energy.

R.; Kapatsina, E.; Sauer, S.; Schreivogel, A.; Tosoni, M. Angew. Chem.,

Int. Ed. 2007, 46, 4832−4887. (c) Sergeyev, S.; Pisula, W.; Geerts, Y.

H. Chem. Soc. Rev. 2007, 36, 1902−1929. (d) Kumar, S. Liq. Cryst.

(14) At the ﬁrst order, U_eff= S²|D| or (S²− 1/4)|D| for integer or

half-integer spin-numbers, respectively.

2009, 36, 607−638. (e) Pisula, W.; Zorn, M.; Young, J. C.; Mullen, K.;

̈

Zentel, R. Macromol. Rapid Commun. 2009, 30, 1179−1202.

(f) Kaafarani, B. R. Chem. Mater. 2011, 23, 378−396.

(32) Terazzi, E.; Bourgogne, C.; Welter, R.; Gallani, J.-L.; Guillon, D.;

Rogez, G.; Donnio, B. Angew. Chem., Int. Ed. 2008, 47, 490−495.

(33) Grumbach, N.; Barla, A.; Joly, L.; Donnio, B.; Rogez, G.;

Terazzi, E.; Kappler, J.-P.; Gallani, J.-L. Eur. Phys. J. B 2010, 73, 103−

108.

(15) Paulsen, C.; Park, J.-G.; Barbara, B.; Sessoli, R.; Caneschi, A. J.

Magn. Magn. Mat. 1995, 140−144, 379−380.

(16) Milios, C. J.; Vinslava, A.; Wernsdorfer, W.; Moggach, S.;

Parsons, S.; Perlepes, S. P.; Christou, G.; Brechin, E. K. J. Am. Chem.

Soc. 2007, 129, 2754−2755.

(17) Soler, M.; Wernsdorfer, W.; Sun, Z.; Ruiz, D.; Huffman, J. C.;

Hendrickson, D. N.; Christou, G. Polyhedron 2003, 22, 1783−1788.

(18) Halcrow, M. A. Chem. Soc. Rev. 2013, 42, 1784−1795.

(19) When the symmetry of the metal complex is lowered from O_hto

D_4hby elongation along the z axis, the original t_2g(d_xy, d_xz, and d_yz)

and e_g(d_z2and d_x2−y2) levels (O_hsymmetry) split into e_g(d_xzand d_yz),

b_2g(d_xy), a_1g(d_z2), and b_1g(d_x2−y2) levels (D_4hsymmetry) according to

the character tables (Figure S31).

(20) Tasiopoulos, A. J.; Vinslava, A.; Wernsdorfer, W.; Abboud, K.

A.; Christou, G. Angew. Chem., Int. Ed. 2004, 43, 2117−2121.

(21) (a) Przychodzen, P.; Pełka, R.; Lewinski, K.; Supel, J.; Rams, M.;

́ ́

Tomala, K.; Sieklucka, B. Inorg. Chem. 2007, 46, 8924−8938. (b) Yang,

C.-I.; Wernsdorfer, W.; Cheng, K.-H.; Nakano, M.; Lee, G.-H.; Tsai,

H.-L. Inorg. Chem. 2008, 47, 10184−10186. (c) Kajiwara, T.; Nakano,

M.; Takaishi, S.; Yamashita, M. Inorg. Chem. 2008, 47, 8604−8606.

(d) Kajiwara, T.; Takahashi, K. I.; Hiraizumi, T.; Takaishi, S.;

̀

(34) Donnio, B.; Riviere, E.; Terazzi, E.; Voirin, E.; Aronica, C.;

Chastanet, G.; Luneau, D.; Rogez, G.; Scheurer, F.; Joly, L.; Kappler,

J.-P.; Gallani, J.-L. Solid State Sci. 2010, 12, 1307−1313.

̀

(35) Riviere, E.; Donnio, B.; Voirin, E.; Rogez, G.; Kappler, J.-P.;

Gallani, J.-L. J. Mater. Chem. 2010, 20, 7165−7168.

(36) Chuard, T.; Deschenaux, R.; Hirsch, A.; Schonberger, H. Chem.

̈

Commun. 1999, 2103−2104.

(37) (a) Felder-Flesch, D.; Rupnicki, L.; Bourgogne, C.; Donnio, B.;

Guillon, D. J. Mater. Chem. 2006, 16, 304−309. (b) Campidelli, S.;

Brandmuller, T.; Hirsch, A.; Saez, I. M.; Goodby, J. W.; Deschenaux,

̈

R. Chem. Commun. 2006, 4282−4284. (c) Gottis, S.; Kopp, C.; Allard,

E.; Deschenaux, R. Helv. Chim. Acta 2007, 90, 957−962.

(38) Terazzi, E.; Jensen, T. B.; Donnio, B.; Buchwalder, K.;

Bourgogne, C.; Rogez, G.; Heinrich, B.; Gallani, J.-L.; Piguet, C.

Dalton Trans. 2011, 40, 12028−12032.

(39) Li, C.-Z.; Matsuo, Y.; Nakamura, E. J. Am. Chem. Soc. 2009, 131,

17058−17059.

Yamashita, M. Polyhedron 2009, 28, 1860−1863. (e) Klower, F.; Lan,

̈

Y.; Nehrkorn, J.; Waldmann, O.; Anson, C. E.; Powell, A. K. Chem.-

Eur. J. 2009, 15, 7413−7422. (f) Inglis, R.; Jones, L. F.; Karotsis, G.;

Collins, A.; Parsons, S.; Perlepes, S. P.; Wernsdorfer, W.; Brechin, E. K.

Chem. Commun. 2008, 5924−5926.

(40) Molard, Y.; Dorson, F.; Cîrcu, V.; Roisnel, T.; Artzner, F.;

Cordier, S. Angew. Chem., Int. Ed. 2010, 49, 3351−3353.

(41) (a) Li, W.; Yi, S.; Wu, Y.; Wu, L. J. Phys. Chem. B 2006, 110,

16961−16966. (b) Yin, S.; Li, W.; Wang, J.; Wu, L. J. Phys. Chem. B

2008, 112, 3983−3988. (c) Li, W.; Yin, S.; Wang, J.; Wu, L. Chem.

Mater. 2008, 20, 514−522. (d) Zhang, T.; Liu, S.; Kurth, D. G.; Faul,

C. F. J. Adv. Funct. Mater. 2009, 19, 642−652. (e) Yang, Y.; Wang, Y.;

Li, H.; Li, W.; Wu, L. Chem.-Eur. J. 2010, 16, 8062−8071. (f) Floquet,

(22) (a) Sieber, A.; Bircher, R.; Waldmann, O.; Carver, G.;

Chaboussant, G.; Mutka, H.; Gudel, H.-U. Angew. Chem., Int. Ed.

̈

2005, 44, 4239−4242. (b) Sanchez-Benitez, A. J.; Kamenev, K. V.;

Moggach, S. A.; Lennie, A. R.; Warren, J. E.; Murrie, M.; Parsons, S.;

Brechin, E. K. Dalton Trans. 2009, 7390−7395. (c) Parois, P.;

Moggach, S. A.; Sanchez-Benitez, J.; Kamenev, K. V.; Lennie, A. R.;

Warren, J. E.; Brechin, E. K.; Parsons, S.; Murrie, M. Chem. Commun.

2010, 1881−1883.

́ ́

S.; Terazzi, E.; Hijazi, A.; Guenee, L.; Piguet, C.; Cadot, E. New J.

Chem. 2012, 36, 865−868.

(42) (a) Nealon, G. L.; Greget, R.; Dominguez, C.; Nagy, Z. T.;

(23) Suzuki, Y.; Takeda, K.; Awaga, K. Phys. Rev. B 2003, 67, 132402.

Guillon, D.; Gallani, J. L.; Donnio, B. Beilstein J. Org. Chem. 2012, 8,

2721

dx.doi.org/10.1021/ja311190a | J. Am. Chem. Soc. 2013, 135, 2708−2722

Journal of the American Chemical Society

Article

349−370. (b) Stamatoiu, O.; Mirzaei, J.; Feng, X.; Hegmann, T. Top.

Curr. Chem. 2012, 318, 331−394.

(43) (a) Kouwer, P. H. J.; Mehl, G. H. J. Am. Chem. Soc. 2003, 125,

11172−11173. (b) Kouwer, P. H. J.; Pourzand, J.; Mehl, G. H. Chem.

Commun. 2004, 66−67. (c) Cardinaels, T.; Driesen, K.; Parac-Vogt, T.

N.; Heinrich, B.; Bourgogne, C.; Guillon, D.; Donnio, B.; Binnemans,

K. Chem. Mater. 2005, 17, 6589−6598.

(44) (a) Donnio, B.; Guillon, D. Adv. Polym. Sci. 2006, 201, 45−155.

(b) Saez, I. M.; Goodby, J. W. Struct. Bonding (Berlin, Ger.) 2008, 128,

1−62. (c) Imrie, C. T.; Henderson, P. A.; Yeap, G.-Y. Liq. Cryst. 2009,

36, 755−777.

(45) For [Mn₁₂O₁₂(L_3,4,5-CB)₁₆(H₂O)₄] at lower temperature, this

diﬀuse scattering is reversibly transformed and substantially sharpened

(the peak position is also slightly shifted to lower angles).

(46) Bourgogne, C.; Bury, I.; Gehringer, L.; Zelcer, A.; Cukiernik, F.;

Terazzi, E.; Donnio, B.; Guillon, D. Molecular Dynamics simulations

of liquid crystalline dendritic architectures. In Atomic-Scale Modeling of

Nanosystems and Nanostructured Materials; Massobrio, C., Bulou, H.,

Goyhenex, Ch., Eds.; Lecture Notes in Physics; Springer: Berlin-

Heidelberg, Germany, 2010; Vol. 795, pp 103−128.

(47) Gatteschi, D.; Sessoli, R. Angew. Chem., Int. Ed. 2003, 42, 268−

297.

(48) (a) Takeda, K.; Awaga, K.; Inabe, T. Phys. Rev. B 1998, 57,

R11062. (b) An, J.; Chen, Z.-D.; Bian, J.; Chen, J.-T.; Wang, S.-X.;

Gao, S.; Xu, G.-X. Inorg. Chim. Acta 2000, 299, 28−34. (c) Aubin, S.

M. J.; Sun, Z.; Eppley, H. J.; Rumberger, E. M.; Guzei, I. A.; Folting,

K.; Gantzel, P. K.; Rheingold, A. L.; Christou, G.; Hendrickson, D. N.

Inorg. Chem. 2001, 40, 2127−2146.

(49) (a) Takeda, K.; Awaga, K. Phys. Rev. B 1997, 56, 14560−14565.

(b) Ruiz-Molina, D.; Gerbier, P.; Rumberger, E.; Amabilino, D. B.;

Guzei, I. A.; Folting, K.; Huffman, J. C.; Rheingold, A.; Christou, G.;

Veciana, J.; Hendrickson, D. N. J. Mater. Chem. 2002, 12, 1152−1161.

(c) Awaga, K.; Suzuki, Y.; Hachisuka, H.; Takeda, K. J. Mater. Chem.

2006, 16, 2516−2521.

(50) (a) Ruiz, D.; Sun, Z.; Albela, B.; Folting, K.; Ribas, J.; Christou,

G.; Hendrickson, D. N. Angew. Chem., Int. Ed. 1998, 37, 300−302.

(b) Zhao, H.; Berlinguette, C. P.; Bacsa, J.; Prosvirin, A. V.; Bera, J.;

Tichy, K.; Schelter, S. E.; Dunbar, E. J. K. R. Inorg. Chem. 2004, 43,

1359−1369. (c) Eppley, H. J.; Tsai, H.; Vries, N.; Folting, K.;

Christou, G.; Hendrickson, D. N. J. Am. Chem. Soc. 1995, 117, 301−

317. (d) Aubin, M. J.; Sun, Z. M.; Eppley, H. J.; Rumberger, R. M.;

Guzei, I. A.; Folting, K.; Gantzel, P. K.; Rheingold, A. L.; Christou, G.;

Hendrickson, D. N. Polyhedron 2001, 20, 1139−1145. (e) Coronado,

́

E.; Forment-Aliaga, A.; Gaita-Arino, A.; Giminez-Saiz, C.; Romero, F.

̃

M.; Wernsdorfer, W. Angew. Chem., Int. Ed. 2004, 43, 6152−6156.

(f) Ruiz-Molina, D.; Mas-Torrent, M.; Gomez, J.; Balana, A. I.;

Domingo, N.; Tejada, J.; Martinez, M. T.; Rovira, C.; Veciana. J. Adv.

Mater. 2003, 15, 42−45.

(51) Four positional isomers have been described in the literature:

1:1:1:1 (S₄symmetry), 2:0:2:0 (D₂symmetry), 1:2:1, and 1:1:2 (the

notation indicates the number of water molecules coordinated to type

II Mn^IIIions).

(52) (a) Wernsdorfer, W.; Ohm, T.; Sangregorio, C.; Sessoli, R.;

Gatteschi, D.; Paulsen, C. Physica B 2000, 284, 1229−1230.

(b) Wernsdorfer, W.; Bhaduri, S.; Tiron, R.; Hendrickson, D. N.;

Christou, G. Phys. Rev. Lett. 2002, 89, 197201. (c) Wernsdorfer, W.;

Allaga-Alcalde, N.; Hendrickson, D. N.; Christou, G. Nature 2002,

416, 406−409.

(53) Artus, P.; Boskovic, C.; Yoo, J.; Streib, W. E.; Brunel, L.-C.;

Hendrickson, D. N.; Christou, G. Inorg. Chem. 2001, 40, 4199−4210.

2722

dx.doi.org/10.1021/ja311190a | J. Am. Chem. Soc. 2013, 135, 2708−2722

Article Doi

Supramolecular organization and magnetic properties of mesogen-hybridized mixed-valent manganese single molecule magnets [MnIII 8MnIV4O12(L x, y, z -CB)16(H2O)4]

DOI: 10.1021/ja311190a

Source and publish data:

Authors:

Article abstract of DOI:10.1021/ja311190a

Full text of DOI:10.1021/ja311190a

Products guided by the article

R&D Labs maybe for 1422746-86-8

Relevant to this article

Hot Product